Quimby Melton, CEO of Confection, sat down with Grit Daily for a long chat about marketing and ad tech, dealing with difficult clients, and the conflicting demands for data privacy and effective online advertising. We learned you can’t make everybody happy, but you can get more bang for your advertising buck than many marketers realize.
Grit Daily: What was your company’s call to action? What inspired you to start confection.io?
Quimby Melton: Before Confection, I ran a marketing agency called Studio Hyperset. Confection’s co-founder and CTO, Bruno Cantuaria, was one of our technology partners. Most of our clients were early-stage SaaS companies in the Bay Area and Seattle, and we helped their marketing operations teams build websites and applications and execute marketing campaigns.
In late 2019, one of our customers emailed us and asked, “Why are our Marketo forms not appearing in Brave?” If you’re unfamiliar with these, Marketo is a CRM owned by Adobe, and Brave is a privacy-focused browser that blocks third-party cookies and cross-domain scripts by default. As we dug into the problem, we realized that Marketo form embeds include Adobe ad and analytics scripts. When Brave blocked the scripts, the form wouldn’t appear.
This is obviously a hard stop from a marketing operational perspective. As we investigated the issue further, we realized blocked analytics scripts and other cross-domain assets also meant website engagement activity – event data – was also going missing. This could include actions like button clicks and pageviews and source information. This loss of this data is catastrophic from a marketing analytics perspective. When pageviews don’t fire and session behavior isn’t recorded, marketers can’t make good decisions about what ads to throttle up or down, which content to emphasize, or which channels are working. There are no reliable signals, just silence.
As we outline here, there’s also a financial component to this as well. Safari and Firefox haven’t supported third-party data since 2017. This means the investments we make in CRMs, DSPs, marketing automation tools, programmatic ads, etc. — they’re only, at most, 75-80% effective. 20-25% of our time and money are effectively wasted, and that’s true every hour of every day. And it’s been compounding daily for five years. In some cases, we have evidence that 89% of events may go missing. And this issue isn’t going away. It’s only going to get more challenging going forward.
Apple recently made privacy-focused changes to the way iOS works. This effectively knocked $200b off Facebook’s market cap. (See this and this) And a Lotame and Business Insider predict the same changes are going to wipe $16 billion in revenue from companies like YouTube, Snap, and Twitter.
Next year, Chrome will begin working more like Safari, Firefox, and iOS. Since 75% of people use Chrome, this accelerates the scale of the challenge facing marketing professionals. Without some intervention, we’ll potentially be wasting the majority of what we spend on ads and marketing software.
As Bruno and I started putting all this together, we realized privacy-first changes are destabilizing a 30-year-old system of moving digital marketing data around. We realized we had an important opportunity to help businesses collect, store, and distribute data in a way that’s unaffected by client-side disruptions involving cookies, cross-domain scripts, and device IDs. And we knew that, as long as we did this in a way that’s compliant with global privacy laws, the product would be good for people too.
Grit Daily: With the rise of big data, are most companies properly handling their user data? Why or why not? And how does Confection.io help companies with this problem?
Quimby Melton: I believe most companies act in good faith and want to strike a balance between compliance and generating the data they need to return value to stakeholders (which, in many cases, is the customer him or herself). They just don’t always know how to do this. There’s a lot of confusion and bad information in this space, and that drives GIGO decision making. In fact, I’m not entirely sure the right mindset exists yet. Digital marketing and advertising are facing a true “BC/AD” moment at present.
Here’s what I mean: If we think back through marketing history, we see a pattern of renting, of treating marketing and advertising as operational expenses (opex). Whether a business ran campaigns in a newspaper or on the radio or on television or via digital ads, businesses effectively paid rent to access distribution networks and audiences they didn’t own.
For example, if a business runs a Super Bowl ad, that business effectively pays to rent (a) the broadcast framework – the licensing, spectrum, cables, antennas, and human resources that make the system run – and (b) the ballgame audience, which the network and sports league invested in, control, and effectively own. We could say the exact same things about print media and radio.
Classic digital marketing, which has been in effect since the 1990s, is the most recent expression of this 100+ year-old model. However, online, not only do we rent distribution frameworks (ad networks, social media, SERP, &c.) and audiences, we also rent our primary ingestion, storage, and distribution mechanisms. This is what a CRM does, for example. Digital marketers own basically nothing, and very few companies control their upstream digital marketing data flows. Instead, they rely on a variety of SaaS products (and those products’ scripts) to ingest, store, and distribute their digital marketing data.
In a world where these ingestion and distribution frameworks begin breaking down because cross-domain scripts fail to fire (or access to cookie information is restricted or depreciated), marketing must move from opex to capex. Marketers must transform themselves from renters to builders and own, at the very least, their upstream data inputs. This “ledger breaking” event is marketing’s true “BC/AD” moment. Things are going to be very different going forward.
In the world that’s coming, marketers will need to build their own self-reliant digital marketing capital assets that allow them to ingest, store, and distribute bulletproof digital marketing data to their critical endpoints (CRMs, applications, analytics systems, &c).
Confection helps companies own their upstream flows, really for the first time. We want to facilitate this transition from marketing-as-renting/opex to marketing-as-owning/capex.
Grit Daily: What kind of data are companies gathering about their users? Does it vary by industry? Why is this data important to these companies?
Quimby Melton: I think there’s a business case for giving people more control over their data. In essence, why would you want to market to people to people who aren’t interested in your product?
However, I also think there’s a consumer case for allowing businesses access to our data. I think popular skepticism arises as a result of some key misunderstandings and naivete. Let me explain:
On the Confection site, we maintain what we call a “not safe for work” index. I use it to log all the heated, aggressive feedback and critiques we receive. My co-founder and I have been threatened, told to commit suicide, and received quite a few (often funny) vulgarities.
Confection is a new kind of product on the far left of the adoption curve so I think of this feedback as the human “immune system” trying to figure out who we are, what we do, and what we stand for. But it’s also evidence of how badly damaged data-focused companies have allowed their relationship with everyday web users to become. Even dipping a toe in the pond unleashes a tidal wave of ferocious blowback.
Let me break the skeptics we encounter into two groups: freethinkers and fanatics. Freethinkers are admirable, rational “show me” personalities. They think deeply, ask questions, and demand proof before acceptance, belief, or adoption. I like these people very much, even if they periodically express themselves in vulgar ways.
Returning to the points we make on the NSFW page, I tell freethinkers what I tell everyone who sends me a colorful message or email handle:
We’re on your side.
Hold me personally responsible for the promises Confection makes.
We care about collecting, storing, and distributing user data in a way that aligns with your personal preferences and complies with data protection laws like CCPA, GDPR, and LGPD. Our customers include businesses you interact with every day, and they feel the same way. Together, we’re working hard to build a better web for you.
Like you, we’re very dissatisfied with the traditional ways data is collected, stored, and distributed online. They’re leaky, inexact, and very intrusive. As such, we rely on a new kind of architecture to generate data and resolve identities.
Confection is quite a bit different from traditional tracking scripts, pixels, and the other things you (understandably) hate. It collects no PII without explicit user consent. It’s compliant with GDPR, CCP, and LGPD on the product level, and users can opt in and out at will.
Our goal is to help businesses thrive in privacy first and to give people better control over what they share online. Our mission is building a new standard for data utility, compliance, and identity management, one that’s good for everyone.
I find these tend to address the concerns of most freethinkers. At least, it gets us to a point where they’ll give us a chance and the benefit of the doubt. However, we also encounter hardline privacy fanatics who see any sort of data tracking (PII or non-PII) as a loss in a binary, zero-sum war they’re waging with marketers, advertisers, and SaaS companies.
Their position is understandable enough. We’ve seen a lot of abuse over the years. However, like all fanatics, people who have these views tend to be binary and myopic. I don’t think they’ve really thought it through.
Would they really exchange a free and open web — supported by ads and data exhaust — for a pay-to-play one? That would create all sorts of problematic hierarchies between users, companies, and governments. What about that book or app or political movement you discovered through more effective targeting? What about all those things you didn’t see because targeting filtered them out?
Fanatics forget that nothing that happens on the web is free. We need financial incentives to offset the overhead costs and financial penalties to establish property lines. If we had an anarchic, completely horizontal web, its usefulness would quickly deteriorate into a true tragedy of the commons.
Undermining solutions like Confection actually limits choice and strengthens entrenched players. By undermining new approaches and solutions that help small- and medium-sized businesses, we’re redistributing all that demand back to monolithic digital players. Hurting solutions like ours helps those players, which is antithetical to (what I assume is) the privacy fanatic’s goal.
We focus on compliant, first-party data sets that establish a direct relationship between our customers and their customers. We don’t share that information or sell it to third parties. Any PII we collect is willfully given. If this doesn’t happen, the user remains anonymous.
Confection lives at the center of a Venn diagram formed by identity resolution, analytics services, privacy services, and lead capture/management. These are sensitive areas, and we understand we’re going to create some friction. We just need to educate people that we have a white-hat goal (better data, more control for people), but we have to adapt what may seem like black-hat methods to give the world what it needs.
Ultimately, this is good piracy at its best, and any freethinker or fanatic who wants to help steer Confection can do so by joining our developer community. That’s a far more constructive action than polluting my demo and threatening my team.
Grit Daily: How does Confection.io go about handling sensitive user data, what makes your approach more secure?
Quimby Melton: Our customers get to choose their compliance framework. Once they select GDPR, CCPA, or LGPD, we only collect PII with a user’s consent. We’ve also engineered the system to not collect passwords or government IDs or HIPAA-sensitive information. We’re solely focused on digital marketing data and events like sources, pageviews, and behavioral interactions.
Also, early on, we made the strategic decision to build our stack using PaaS vendors rather than some sort of bare metal solution. By leveraging Heroku and Snowflake, we benefit from world-class security and redundancy protocols. We’ve never regretted this decision.
When people think of user data collection/use, typically what comes to mind are tech and social media companies, what other industries would you say modern data practices have impacted the most? What industries/companies does Confection.io target? Has the target changed since you started?
Quimby Melton: Martech and adtech are famously large addressable markets. From the beginning, we knew our target customer was anyone with a site or app impacted by the disruptions we’ve discussed. The issue is that widespread.
In our pitch deck, we use the story of Sam to ground this a bit.
Sam is a digital marketer. Like every person and business that operates a website, he’s wasting at least 20% of every $1 he spends on digital ads and marketing software. Sam could work in a corporate marketing department, at a publisher, or at an agency. He could be an application developer or run an independent ad network. The changes we’ve discussed will impact him on the marketing side and, if he has one, the product side as well.
For example, before he installed Confection, Sam didn’t know how many pageviews his site was actually getting. As a result, he couldn’t set his ad rates correctly. Recovering lost pageviews helped him make $2,000 more per month in ad revenue.
Sam also didn’t know how effective his paid marketing channels were. By turning off poor performers and doubling down on high performing ones, Sam is able to make far better decisions about how he spends his marketing budget.
Both of these examples illustrate real results from actual Confection users, and I think they show that data collection and usage impacts more than just social media sites. In fact, I’d say B2B interests – which most people are probably less familiar with – are impacted much, much more than B2C interests. ABM strategies live and die on behavioral data and knowledge about individual companies and team members. B2C interests can always return to a mass media model. B2B marketers – most of whom sell in narrow verticals at high price points – can’t do that as easily.